Download Cis-elements of protein transport to the plant vacuoles

Journal of Experimental Botany, Vol. 50, No. 331, pp. 165–174, February 1999
Cis-elements of protein transport to the plant vacuoles
Ken Matsuoka1 and Jean-Marc Neuhaus2,3
1Laboratory of Biochemistry, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa-ku,
Nagoya 464–8601, Japan
2Laboratoire de Biochimie, Université de Neuchâtel, rue Émile-Argand 9, CH-2007 Neuchâtel 7, Switzerland
Received 3 March 1998; Accepted 16 July 1998
Abstract
Vacuolar proteins are synthesized and translocated
into the endoplasmic reticulum and transported to the
vacuoles through the secretory pathway. Three different types of vacuolar sorting signals have been identified, carried by N- or C-terminal propeptides or internal
sequences. These signals are needed to target proteins to the different types of vacuoles that can coexist
in a single plant cell. A conserved motif (NPIXL or
NPIR) was identified within N-terminal propeptides, but
can also function in a C-terminal propeptide and targets proteins in a receptor-mediated manner to a lytic
vacuole. Binding to a family of putative sorting receptors for sequence-specific vacuolar sorting signals has
been used as an assay to identify further peptides with
other binding motifs. No motif was found in C-terminal
sorting sequences, which need an accessible terminus, suggesting that they are recognized from the
end by a still unknown receptor. The phosphatidylinositol kinase inhibitor wortmannin differentially affects
sorting mediated by these two sorting sequences, suggesting different sorting mechanisms. Less is known
about sorting mediated by internal protein sequences,
which do not contain the conserved motif identified in
N-terminal propeptides and may function by aggregation, leading to transport by coat-less dense vesicles
to protein storage vacuoles. Even less is known about
the sorting of tonoplast proteins, for which several
sorting systems will also be needed.
Key words: Protein trafficking, propeptides, vacuole, tonoplast, sorting receptor.
Introduction
The plant cells possess one or several vacuoles that fulfil
various functions. Transport of solutes through the tono-
plast is required for the turgor pressure. Vacuoles store
proteins, polysaccharides, organic acids, and pigments.
Detoxified compounds are deposited in vacuoles, but also
precursors of toxic compounds needed for defence ( Wink,
1993). The vacuoles belong to the secretory pathway, since
they are derived from the endoplasmic reticulum (ER),
where both its soluble and tonoplast proteins are synthesized. The soluble proteins of the secretory pathway require
a signal sequence for synthesis by membrane-bound ribosomes on the ER. After the protein entered the ER, its
signal sequence is removed and the protein is released into
the ER lumen. Membrane proteins are similarly inserted
into the ER membrane. These proteins are exported from
the ER and targeted to their final destinations by vesicular
transport. The final location depends on targeting information within the polypeptides (Höfte et al., 1991). This review
will focus on the structural determinants that are required
for the sorting of vacuolar protein precursors and for their
transport to the final compartment. The identification of
three different types of sorting signals will be linked to the
recent visualization of different vacuolar compartments
within a single plant cell and to the differential effect of the
lipid kinase inhibitor, wortmannin, on the sorting of two
different vacuolar proteins. There will be a short discussion
of tonoplast protein sorting, for which there is only scant
information.
Targeting of soluble proteins to a vacuole requires
specific peptide information
For the secretory pathways of animals the concept of a
default pathway of secretion has been proposed.
According to this ‘bulk-flow’ model, any soluble protein
cotranslationally transported into the ER will be carried
passively to the Golgi and from there to the surface of
the cell and will eventually be released into the extracellu-
3 To whom correspondence should be addressed. Fax +41 32 718 22 01. E-mail: [email protected]
Abbreviations: CTPP, C-terminal propeptide; ER, endoplasmic reticulum; NTPP, N-terminal propeptide; VSS, vacuolar sorting sequence.
© Oxford University Press 1999
166
Matsuoka and Neuhaus
lar medium unless it carries sorting signals (Rothman
and Wieland, 1996). These sorting signals will cause a
receptor-mediated accumulation in transport carrier vesicles. This model applies both to the lysosomal targeting
by mannose-6-phosphate receptors and to the recycling
of escaped ER-resident proteins. In the ER itself, sorting
signals may contribute both to the specific concentration
of secreted proteins in COPII vesicles and cause the
exclusion of ER-resident proteins from such carriers
(Balch and Farquhar, 1995; Bannykh and Balch, 1997).
Essentially, every transport step involving the formation
of specific (coated) vesicles can be expected to cause the
concentration of some proteins and the exclusion of
others. The concept of a default pathway is more suitable
for transport steps involving larger compartments, as in
the cisternal maturation model of Golgi transport. This
model supported by the direct observation of scale maturation in algae (Becker and Melkonian, 1996), and also
the visualization of protein-GFP transport in animal cells
(Lippincott-Schwartz and Smith, 1997). There is also a
default pathway at single branch points, where a dominant sorting signal determines the fate of a protein. This
is the case for the lysosomal sorting of soluble proteins
carrying a mannose-6-phosphate in animal cells and for
soluble vacuolar proteins carrying a specific propeptide
in yeast (Horazdovsky et al., 1995). It also applies to
plant vacuolar proteins for which such dominant signals
have been identified.
Several plant vacuolar proteins have been analysed and
in most cases specific determinants were necessary for
proper targeting. When these determinants were deleted
or mutated, the proteins were secreted into the extracellular space. The determinants could be fused to secreted
proteins, transforming them into vacuolar proteins, confirming that secretion is a default pathway for soluble
proteins and that vacuolar sorting signals are dominant
determinants of protein localization. In all cases analysed
the vacuolar sorting signal was found in the polypeptide
sequence itself. Glycan side-chains did not contribute to
the targeting, in contrast to lysosomal proteins in mammalian cells (Sonnewald et al., 1990; Wilkins et al., 1990).
Three different types of vacuolar sorting signals
(VSS) have been identified
Many of the vacuolar sorting signals were found in Nor C-terminal propeptides, but not all propeptides of
vacuolar proteins contain targeting information. In this
case, or in vacuolar proteins lacking propeptides, the
targeting information must be present within the mature
polypeptide chain.
The C-terminal targeting propeptides of chitinases,
glucanases or osmotins were identified by comparison of
predicted protein sequences with those of related secreted
proteins (Neuhaus et al., 1991; Sticher et al., 1992;
Melchers et al., 1993). The C-terminal targeting propeptides of cereal lectins and the N-terminal propeptide of
sporamin were identified by comparison of the precursor
polypeptide with the mature protein (Matsuoka and
Nakamura, 1991; Dombrowski et al., 1993; Saalbach
et al., 1996). The N-terminal targeting propeptide of
aleurain was identified by comparison with the very
different propeptide of a related secreted protease
(Holwerda et al., 1992). One internal targeting sequence
was tentatively located in bean phytohaemagglutinin by
fusion to invertase of a series of truncated polypeptides
(von Schaewen and Chrispeels, 1993).
The sequence requirement for the vacuolar sorting
signals found in the propeptides were determined by
deletion and mutation, fusion to reporter proteins and
comparison of the sequences. Some of these signals were
also characterized by other criteria, such as the affinity
to a putative sorting receptor or the sensitivity of targeting
to the compound wortmannin in tobacco BY-2 cells.
Based on these results, it is possible to distinguish three
types of vacuolar sorting signals, corresponding probably
to three different sorting systems. (1) Sequence-specific
VSS (ssVSS) can be located in N- or C-terminal propeptides as well as within mature proteins, contain conserved
motifs and bind to one or several members of the putative
vacuolar sorting receptors ( VSR). Their sorting occurs
by means of clathrin-coated vesicles and is insensitive to
wortmannin. (2) C-terminal VSS (ctVSS ) with low
sequence specificity must be located at the very end of
the polypeptide. Their sorting is sensitive to wortmannin.
(3) Protein structure-dependent VSS (psVSS) may be
determinants of aggregation and mediate transport to
vacuoles by means of dense vesicles.
Sequence-specific vacuolar sorting signals
(ssVSS)
To date, sorting signals have been extensively studied in
the N-terminal propeptides (NTPPs) of sweet potato
sporamin and barley aleurain (Matsuoka and Nakamura,
1991; Holwerda et al., 1992). These NTPPs contain a
common sequence NPIRL/P, which can also be found in
the NTPPs of proteins related to either of these two
proteins, the Kunitz-type protease inhibitors of potato
tuber (Ishikawa et al., 1994) and the family of plant
vacuolar proteases, respectively ( Fig. 1). Most potato
protease inhibitors also have a similar motif NPLDV
close to their C-terminus.
Although the deletion analysis indicated that the region
of the NPIRL/P motif is critical for its function
(Holwerda et al., 1992; Nakamura et al., 1993), a strict
conservation of these amino acids may not be essential
to constitute the vacuolar sorting signal. Detailed mutation analysis of this region in sporamin was carried out
by replacing each of the five amino acids of the motif by
Cis-elements of protein transport
167
Fig. 1. Comparison of sequence-specific vacuolar sorting sequences (ssVSS ) within precursors of vacuolar proteins. Sequences identified by binding
in vitro to a BP-80 protein are underlined. The conserved NPIXL or NPIR motif is indicated in bold. The conserved NLPS motif is indicated in
italics. Processing sites of the propeptides are indicated by |.
at least ten different amino acids and expressing the
substitution mutants in tobacco BY-2 cells ( K Matsuoka
and K Nakamura, unpublished results). Based on this
analysis, the following properties of amino acids are
required to constitute the VSS: the first amino acid (N in
sporamin) should not be small and hydrophobic (A or
V ). The second position (P) may not be acidic, since E
and D cannot substitute for P without decreasing the
sorting efficiency. The large and hydrophobic alkyl chain
of the third position (I ) is essential. This isoleucine can
only be changed to leucine without losing the VSS function. Substitution with V or M decreased the sorting
efficiency by about 50%, and F or any other tested amino
acids caused almost complete secretion of sporamin. The
fourth position (R) can be any amino acid, and this
amino acid is not conserved in the NTPPs of potato
protease inhibitors. The fifth position (L in the case of
sporamin and potato protease inhibitors and P in aleurain) should contain a bulky and preferably hydrophobic
side-chain, because the exchange with G, A or S abolished
vacuolar targeting completely. Thus, many sequences that
do not fit the strict NPIRL/P sequence can function as
vacuolar sorting signals, although the signal should contain one core L or I. In aleurain, the sorting efficiency
depends not only on the NPIRP sequence, but also on
whole or part of the upstream and downstream flanking
sequences (Fig. 1; Holwerda and Rogers, 1993).
The identification of a putative vacuolar sorting recep-
tor of 80 kDa (BP-80) by affinity chromatography on a
column with a coupled aleurain VSS peptide ( Kirsch
et al., 1994) allowed the testing of potential VSS-containing peptides for their affinity to this sorting system,
which were comparable to the results discussed above for
sporamin. Binding of an 80 kDa protein from the same
pea extract strongly enriched in clathrin-coated vesicles
was the criterion ( Kirsch et al., 1996). In this system,
binding was found with the VSS of aleurain but not with
the propeptide of the related secreted endopeptidase. The
propeptide of sporamin also bound, while the propeptide
of barley lectin did not bind, as expected. Single and
double replacements confirmed the importance of Ile
within the VSS of sporamin. Its replacement by a Gly
abolished binding, while substitution by a Met had no
effect. Another 22 residue long peptide, derived from the
C-terminus of the precursor of the 2S albumin from
Brazil nut (Saalbach et al., 1996), was also found to bind
a BP-80. Progressive deletions from the N-terminus
caused progressive loss, but then recovery of a BP-80
binding for an 11 amino acids long peptide containing
the terminal nine amino acids. These nine terminal amino
acids did not contain a NPIRP/L sequence, though they
contained a ProMet and an Ile that could be involved in
binding to the receptor. On the other hand, deletion of
the four last amino acids (IAGF ), which corresponded
to the propeptide of this 2S albumin, abolished binding.
This propeptide has been found to be necessary, though
168
Matsuoka and Neuhaus
not sufficient, for vacuolar targeting in vivo (Saalbach
et al., 1996). It is also not excluded that different peptides
bind different proteins of similar molecular weight, e.g.
different isoforms of BP-80 that are known from different
cDNA clones (Paris et al., 1997). Preliminary results
seem to indicate that one peptide can deplete the extract
from the protein that would bind another unrelated
peptide. The peptides may also have widely different
affinities not detected in this assay system, but at least
the aleurain peptide was found to have a much higher
affinity for BP-80 than the sporamin peptide ( Kirsch
et al., 1994).
In a similar work with extracts of vesicles from pumpkin, binding of a 72 and an 82 kDa protein (PV72 and
PV82) was observed to the aleurain peptide as well as to
two peptides derived from the pro-2S albumin from
pumpkin itself (Shimada et al., 1997). One peptide
included an internal propeptide while the other included
the C-terminal region of the precursor. Binding to the
internal peptide was traced down to a region containing
a processing site (MRGIEN|PWRREG), and a glycine
scan identified two important residues, the R and E
indicated in bold. Since this region is exposed on the
surface of the albumin precursor, as evidenced by the
processing site, it is quite possible that it is a VSS in vivo.
The C-terminal peptide ( KARNLPSMCGIRPQRCDF )
contains an NLPS motif conserved in 2S albumins of
other species (at the homologous position in castor bean
and Brazil nut, but positioned further C-terminally in
Arabidopsis). Replacement of the four residues by GGGG
abolished the binding. Thus NLPS could still be another
motif recognized by proteins of the BP-80 family (or
VSR, Paris et al., 1997). It is also present in the
N-terminal propeptides of several potato proteinase inhibitors related to sporamin, linked to the NPI sequence
(NPINLPS, Ishikawa et al., 1994). It could also be an
indication that the receptor can bind I/L-containing peptides in both orientations. Again, as for the Brazil nut
albumin, the binding affinity of the pumpkin 2S albumin
peptide is unknown and the relevance of this result is
unclear as long as the motifs have not been shown to be
involved in vacuolar sorting of complete proteins in vivo.
On the other hand, a peptide representing the 17
C-terminal amino acids of the pro-2S albumin from
Arabidopsis, including a motif NIPS, was reported not to
have bound a BP-80 ( Kirsch et al., 1996).
While the mutation analysis only detected one essential
amino acid, I/L, the comparison of an increasing number
of sequence-specific VSSs will probably allow the identification of more consensus positions. While many random
sequences could replace a signal sequence or a mitochondrial transit peptide, it was still possible to derive consensus rules to identify natural signal or transit peptides
(Baker and Schatz, 1987; Kaiser et al., 1987; von Heijne,
1990). There is also obviously no need for sequence-
specific VSSs to be included within an N-terminal propeptide. The targeting of sporamin to the vacuole was correct
even when the N-terminal propeptide was moved to a
C-terminal position ( Koide et al., 1997). Some proteins
without propeptides may thus turn out to harbour their
VSS within the mature polypeptide.
Transport to the vacuole of proteins with a ssVSS was
found to be relatively resistant to wortmannin in tobacco
BY-2 cells, compared with another group of vacuolar
proteins described below (Matsuoka et al., 1995).
C-terminal vacuolar sorting signals (ctVSS)
Vacuolar proteins with a C-terminal propeptide were
secreted when tobacco BY-2 cells were treated with wortmannin. This indicates a biochemical difference between
two sorting systems and may be used as a diagnostic test
for the type of VSS. However, the differential sensitivity
to wortmannin of ssVSS and ctVSS can be observed in
BY-2 cells, but not in many other plant cells where
wortmannin caused secretion of both protein groups at a
lower concentration (1 mM instead of 10–33 mM ). The
different levels of phosphatidylinositol (PI ) 3-kinase activity may explain this difference. In BY-2 cells at log phase,
PI 3-kinase activity was very high and the synthesis of PI
3-phosphate was not inhibited completely with relatively
high concentrations of wortmannin (Matsuoka et al.,
1995). By contrast, many plant cells and tissues express
very low levels of PI 3-kinase activity and most of these
activities are inhibited at lower concentrations of wortmannin ( K Matsuoka, unpublished result).
By contrast with the sequence-specific VSSs, no
common motif was found in the vacuolar sorting signals
identified in the C-terminal propeptides of chitinases or
cereal lectins. Several different segments of the barley
lectin propeptide were each functional in vivo. A minimum
length of three (!) residues still caused predominantly
vacuolar localization of the lectin in tobacco and four
Ala were as effective as LLVD or PIRP, but four Glu,
four Lys or four Gly were ineffective (Dombrowski et al.,
1993). Similarly, in the tobacco chitinase A, six residues
were sufficient, many single replacements had little effect
and several random sequences could also function as
vacuolar sorting signals (Neuhaus et al., 1994).
Importantly, rendering the sequence more hydrophilic or
more hydrophobic did not strongly affect the sorting
efficiency.
In both cases, the function of these vacuolar sorting
signals could be most efficiently reduced by the addition
of one or several Gly residues or of an N-glycosylation
site to the end of the propeptide. This means that the
ctVSS must be accessible from the end, similar to the ER
retention signals -HDEL or -KDEL or the peroxisomal
targeting signal -SKL, which can also be blocked by the
addition of amino acids.
Cis-elements of protein transport
Sequence comparison of natural sequences may be
more informative than analysis of mutants. As mentioned
above, mutation analysis is a less sensitive method to
identify consensus motifs than comparison of natural
sequences. Therefore sequences of C-terminal propeptides
were collected from the protein families for which at least
one member had been shown to have a C-terminal VSS
( Fig. 2). Comparison indicates a preference for Met at
the last position, along with Leu, Ile and Val. Amino
acids with a long aliphatic side-chain may also carry a
terminal charge, as in Glu or Lys. The next two amino
acids are most often hydrophilic and/or negatively
charged. There is a slight preference for hydrophobic
side-chains at the next more distal positions. There may
also be species differences as suggested by the divergence
of both chitinase and glucanase propeptides in bean, both
due to a frame shift. Cereals also may have different
preferences.
From these comparisons, the hypothetical receptor is
expected to bind the propeptide at its terminal carboxyl
group by interactions involving the backbone of the
peptide. To encompass the C-terminal amino acid there
must be a hydrophobic patch requiring at least a methyl
group, but preferably two or three methylene groups,
while the end group of the side-chain may not bind at
all. Such binding pockets have been described in animal
proteins (Saras and Heldin, 1996).
C-terminal extensions or truncations can arise easily
by frameshift or point mutations without affecting the
proper folding and the activity of proteins. Thanks to the
low sequence specificity, secreted and vacuolar proteins
can easily be reoriented to the other compartment.
Vacuolar proteins using this sorting system are not digestive enzymes (at least not proteases) and are easily tolerated
in the extracellular space, where related isoforms are often
already present (e.g. chitinases, glucanases and other
pathogenesis-related proteins).
The ER retention signal H/KDEL resembles a ctVSS
and could also be able to mediate vacuolar sorting.
Indeed, when overexpressed, a portion of sporamin with
an HDEL extension escaped from the ER and was
recovered in the vacuole (Gomord et al., 1997). This may
enable the plant cell to recycle escaped ER proteins but
it may also be a mechanism controlling the transport of
certain proteins to a non-lytic vacuole. An example could
be the cysteine endoproteinase from mung bean, which is
related to aleurain but harbours a terminal -KDEL and
degrades seed globulins after germination (Okamoto
et al., 1994). In the same protease families, there are also
proteases with an additional Cys-rich C-terminal domain
which ends with what could well be a ctVSS, e.g. -LIDNM
in oryzain b.
Some C-terminal propeptides may combine both sorting signals described so far. In the case of the Brazil nut
2S albumin, the C-terminal 9 amino acids (including the
169
Fig. 2. Sequence comparison of C-terminal propeptides of several
protein families. Glycosylation sites are indicated in italics, negative
charges near the C-terminus are indicated in bold. Processing sites of
some propeptides identified by terminal sequencing of the mature
proteins are indicated by |, for other propeptides they were placed by
comparison with secreted members of the same protein family and are
suggested by the spacing.
four amino acids long propeptide) constitute the vacuolar
targeting signal (Saalbach et al., 1991) and bind to a
BP-80 ( Kirsch et al., 1996). It is possible that the propeptide of the tobacco osmotin AP-24 uses both mechanisms,
as it contains a sequence FPLEM that fits to the minimum
sequence requirement defined with sporamin (see above).
As an artificial C-terminal propeptide, the N-terminal
170
Matsuoka and Neuhaus
propeptide of prosporamin functioned as a vacuolar
sorting signal, but targeting was still wortmannininsensitive ( Koide et al., 1997). Replacement of the
conserved Ile by Gly in this displaced propeptide did not
cause secretion as it did in the original position, but
rendered the vacuolar sorting wortmannin-sensitive as for
a bona fide ctVSS.
The failure so far to identify a sorting receptor for
ctVSSs could be a consequence of their low sequence
specificity. It could, however, also be due to a receptorfree sorting mediated by aggregation, as discussed below
for the third type of VSS.
Physical structure vacuolar sorting signals
(psVSS)
Some storage proteins are thought to have a sorting
signal within the mature polypeptide chain. This may be
due to lack of information on processing, but several seed
proteins are sorted by a different mechanism. Seed proteins of the vicilin and legumin families accumulate in
dense vesicles that bud off from the trans-Golgi without
the participation of clathrin coats and accumulate in
protein storage vacuoles ( Robinson et al., 1998).
Aggregation has been proposed as a non-receptormediated sorting mechanism ( Vitale and Chrispeels, 1992)
and is known to occur in a pH-dependent manner in
animal cells. If aggregation occurs rapidly after synthesis,
protein bodies can form from the ER as is the case for
cereal prolamins (Levanony et al., 1992). In legumes,
oligomerization is an important determinant of transport
competence ( Vitale et al., 1995) and affects the accessibility of hydrophobic patches on the surface of the protein.
Pea prolegumin is more hydrophobic in the ER and Golgi
than is legumin in the protein bodies (Hinz et al., 1997).
Processing in the Golgi may further uncover aggregation
determinants. With such a mechanism, it is difficult to
locate the sorting signals precisely.
One such internal signal was tentatively located in a
surface loop of bean phytohaemagglutinin (PHA) by the
use of fusions of truncated PHA to invertase (von
Schaewen and Chrispeels, 1993). This internal loop could
contain a ssVSS (GFLGL could fit). However, truncation
may also have exposed the surface sequences that cause
aggregation without necessarily playing any role in the
sorting of the intact protein precursor. The role of the
C-terminal propeptide of PHA has also not yet been
addressed ( Young et al., 1995), as it could function either
as a ctVSS or by modulating the exposition of an internal
aggregation determinant. It is not excluded that BP-80
proteins could also play a role in transporting malfolded
proteins to the lytic vacuole if they escaped the quality
control in the ER, as does the CPY receptor Vps10p for
malfolded proteins in yeast (Hong et al., 1996).
Two or three sorting mechanisms, two or three
target vacuoles?
The sorting system for sequence-specific VSSs has been
compared with and is clearly distinct from the other two
sorting systems, but the latter two have not been clearly
compared yet. It is known that ssVSS-carrying proteins
are exported from the Golgi apparatus to lytic vacuoles
by clathrin-coated vesicles while psVSS-containing proteins are exported to protein bodies by apparently coatless dense vesicles (Robinson et al., 1998). It is also
known that proteins with a ctVSS are sorted to a vacuole
distinct from the lytic vacuole (Paris et al., 1996) by a
wortmannin-sensitive mechanism (Matsuoka et al., 1995).
This could even be seen in live cells by the use of a Green
Fluorescent Protein (GFP) fused to the C-terminal VSS
of tobacco chitinase. This GFP accumulated in the large
central vacuole of chloroplast-rich mesophyll protoplasts,
but mostly in a small vacuole in chloroplast-poor protoplasts, while the large vacuole remained intact. Staining
with the acid-trapped dye Neutral Red indicated that the
fluorescent vacuoles had a neutral pH while the nonfluorescent vacuoles were acidic. One vacuole type was
large and the other small, depending on the cell type (Di
Sansebastiano et al., 1998). On the other hand, there is
little information about the vesicles transporting proteins
with a ctVSS and little idea of the effect of wortmannin
on the transport of aggregated proteins by dense vesicles.
It is thus unclear whether there are two or three distinct
transport pathways to two or three distinct vacuole types.
This is further complicated by the possibilities of convergence of pathways to a single vacuole and of fusion of
different vacuole types that may explain the finding of
sporamin and barley lectin in aggregates in the same
central vacuole (Schroeder et al., 1993).
One protein may well be transported to two different
vacuoles, such as the barley aspartic proteinase that is
found both in protein storage vacuoles and in lytic
vacuoles (Paris et al., 1996). Its precursor contains both
an NPIR-containing propeptide and an internal insertion
of a saposin-like insert that was found to be involved in
vacuolar targeting ( Törmäkangas, 1997). The sporamin
with a C-terminally transplanted propeptide also indicates
this possibility ( Koide et al., 1997). It would be interesting
to know whether it is transported to two different vacuoles
or whether the VSR-dependent sorting is dominant. There
is also the more general question of the spatial and
temporal order of action of these sorting systems within
the Golgi. A sorting system acting earlier (in a more cis
compartment) would prevent sorting by a later-acting
system.
Processing of targeting propeptides
Many vacuolar proteins are processed after their sorting.
In many cases the processing would occur after arrival in
Cis-elements of protein transport
the vacuole. Vacuolar endoproteases involved in processing have been identified in several plants (HaraNishimura et al., 1991). One of these enzymes preferentially cuts Asn-Gly bonds (Scott et al., 1992), a bond
frequently found in precursors of seed storage proteins,
but also in tobacco chitinase and AP-24 (Fig. 2). This
site-specific processing may be followed by carboxypeptidase trimming, which may lead to ragged ends (Nagahora
et al., 1992). It was found that mutations of the Asn-Gly
site did not affect processing of the chitinase, which could
be processed exclusively by a carboxypeptidase (Freydl,
1995).
In contrast to the processing of seed storage proteins,
tobacco chitinase and AP-24, the processing of NTPPs
during the transport are more complex in both sporamin
and aleurain, as intermediate forms were found during
the maturation of these proteins (Holwerda et al., 1992;
Nakamura et al., 1993). The presence of these intermediate precursors raises the question, whether all proteolytic cleavages occur in the vacuole. One possibility is
that the processing of the most N-terminal part of the
propeptide, usually containing the ssVSS, occurs just after
the transport from the Golgi apparatus to the newly
identified intermediate compartment between the Golgi
apparatus and vacuole, the ‘prevacuole’ (Paris et al.,
1997) or ‘Pep12 compartment’ (da Silva Conceição et al.,
1997). Further processing, such as the activation of
the protease aleurain by the removal of the bulk of the
N-terminal propeptide, a protease inhibitor domain, could
occur after arrival to the vacuole from this intermediate
compartment.
171
Golgi of a chimaeric VSR protein (Jiang and Rogers,
1998). Differential sensitivity to inhibitors like monensin
or brefeldin A also hinted at different sorting pathways
for soluble and membrane proteins (Gomez and
Chrispeels, 1993). The transport of TIP may thus involve
a completely different pathway than for soluble proteins,
a pathway avoiding altogether the Golgi apparatus. The
extreme abundance of TIPs in tonoplasts suggests in fact
a structural role, with each TIP subtype defining a different vacuolar compartment (discussed in Neuhaus and
Rogers, 1998).
Conclusion
Protein transport and sorting in plant cells remains a field
with many open questions. The fundamental mechanisms
can be expected to be similar to other eukaryotes, as
evidenced by the identification of homologues to proteins
involved in protein trafficking in the Expressed Sequence
Tag ( EST ) programs. However, sorting to the plant
vacuole has specific features that need further clarification. The existence of several different vacuolar compartments, several different vacuolar sorting signals and the
different sorting pathways postulated for tonoplast proteins all indicate that with plants there is much more to
do than confirming and extending models developed for
other systems. One family of sorting receptors has been
identified, but it may only sort a fraction of the vacuolar
proteins. One or more additional receptors still await
identification. We are therefore looking forward to exciting new results that will clarify the biogenesis of the
vacuoles and the function of these important organelles.
Sorting of tonoplast proteins
Analysis of targeting of tonoplast proteins has been much
more difficult than for the soluble vacuolar proteins. This
may be due to the lack of known simple type I or type
II proteins that made many advances possible in animal
systems. The most abundant tonoplast proteins are integral membrane proteins ( TIPs) with six transmembrane
segments and both N- and C-termini on the cytosolic
face of the tonoplast. The different TIP isoforms are
located in different vacuoles in pea and barley root cells
and may thus be sorted by different mechanisms (Paris
et al., 1996). As for the soluble proteins, they can also
be found in one and the same vacuole. The C-terminal
transmembrane and cytosolic tail of a-TIP was found to
be sufficient to target a reporter protein to the tonoplast,
and truncation of the cytosolic tail did not affect this
result (Höfte and Chrispeels, 1992). This could mean that
the transmembrane segment alone carries the sorting
information or that a tonoplast (of which vacuole type?)
is the default destination for membrane proteins as found
for yeast (Roberts et al., 1992; Wilcox et al., 1992). The
C-terminal tail of a-TIP prevented transit through the
Acknowledgements
The authors would like to Nadine Paris for useful discussions.
The work described here is supported by the Swiss National
Science Foundation, Grant No. 31–46926.96 and by grants
from the Ministry of Education, Science and Culture, Japan.
References
Baker A, Schatz G. 1987. Sequences from a prokaryotic genome
or the mouse dihydrofolate reductase gene can restore the
import of a truncated precursor protein into yeast mitochondria. Proceedings of the National Academy of Sciences, USA
84, 3117–3121.
Balch WE, Farquhar GM. 1995. Beyond bulk flow. Trends in
Cell Biology 5, 16–19.
Bannykh SI, Balch WE. 1997. Membrane dynamics at the
endoplasmic reticulum-Golgi interface. Journal of Cell Biology
138, 1–4.
Becker B, Melkonian M. 1996. The secretory pathway of
protists—spatial and functional organization and evolution.
Microbiological Reviews 60, 697 ff.
da Silva Conceição A, Marty-Mazars D, Bassham DC,
Sanderfoot AA, Marty F, Raikhel NV. 1997. The syntaxin
172
Matsuoka and Neuhaus
homolog AtPEP12p resides on a late post-Golgi compartment
in plants. The Plant Cell 9, 571–582.
Di Sansebastiano GP, Paris N, Marc-Martin S, Neuhaus J-M.
1998. Specific accumulation of GFP in a non-acidic vacuolar
compartment via a C-terminal propeptide-mediated sorting
pathway. The Plant Journal 15, 449–457.
Dombrowski JE, Schroeder MR, Bednarek SY, Raikhel NV.
1993. Determination of the functional elements within the
vacuolar targeting signal of barley lectin. The Plant Cell
5, 587–596.
Freydl E. 1995. Analysis of the targeting peptide of chitinase.
PhD thesis, University of Basel.
Gomez L, Chrispeels MJ. 1993. Tonoplast and soluble vacuolar
proteins are targeted by different mechanisms. The Plant Cell
5, 1113–1124.
Gomord V, Denmat LA, Fitchette-Lainé AC, Satiat-Jeunemaitre
B, Hawes C, Faye L. 1997. The C-terminal HDEL sequence
is sufficient for retention of secretory proteins in the
endoplasmic reticulum ( ER) but promotes vacuolar targeting
of proteins that escape the ER. The Plant Journal 11, 313–325.
Hara-Nishimura I, Inoue K, Nishimura M. 1991. A unique
vacuolar processing enzyme responsible for conversion of
several proprotein precursors into the mature forms. FEBS
Letters 294, 89–93.
Hinz G, Menze A, Hohl I, Vaux D. 1997. Isolation of prolegumin
from developing pea seeds—its binding to endomembranes
and assembly into prolegumin hexamers in the protein storage
vacuole. Journal of Experimental Botany 48, 139–149.
Höfte H, Chrispeels MJ. 1992. Protein sorting to the vacuolar
membrane. The Plant Cell 4, 995–1004.
Höfte H, Faye L, Dickinson C, Herman EM, Chrispeels
MJ. 1991. The protein-body proteins phytohemagglutinin
and tonoplast intrinsic protein are targeted to vacuoles in
leaves of transgenic tobacco. Planta 184, 431–437.
Holwerda BC, Padgett HS, Rogers JC. 1992. Proaleurain
vacuolar targeting is mediated by short contiguous peptide
interactions. The Plant Cell 4, 307–318.
Holwerda BC, Rogers JC. 1993. Structure, functional properties
and vacuolar targeting of the barley thiol protease, aleurain.
Journal of Experimental Botany 44, 321–329.
Hong E, Davidson AR, Kaiser CA. 1996. A pathway for
targeting soluble misfolded proteins to the yeast vacuole.
Journal of Cell Biology 135, 623–633.
Horazdovsky BF, DeWald DB, Emr SD. 1995. Protein transport
to the yeast vacuole. Current Opinion in Cell Biology
7, 544–551.
Ishikawa A, Ohta S, Matsuoka K, Hattori T, Nakamura K.
1994. A family of potato genes that encode Kunitz-type
proteinase inhibitors: structural comparisons and differential
expression. Plant Cell Physiology 35, 303–312.
Jiang L, Rogers JC. 1998. Integral membrane protein sorting
to vacuoles in plant cells: evidence for two pathways. Journal
of Cell Biology 143, 1183–1199.
Kaiser CA, Preuss D, Grisafi P, Botstein D. 1987. Many random
sequences functionally replace the secretion signal sequence
of yeast invertase. Science 235, 312–317.
Kirsch T, Paris N, Butler JM, Beevers L, Rogers JC. 1994.
Purification and initial characterization of a potential plant
vacuolar targeting receptor. Proceedings of the National
Academy of Sciences, USA 91, 3403–3407.
Kirsch T, Saalbach G, Raikhel NV, Beevers L. 1996. Interaction
of a potential vacuolar targeting receptor with amino- and
carboxyl-terminal targeting determinants. Plant Physiology
111, 469–474.
Koide Y, Hirano H, Matsuoka K, Nakamura K. 1997. The
N-terminal propeptide of the precursor to sporamin acts as
a vacuole-targeting signal even at the C-terminus of the
mature part in tobacco cells. Plant Physiology 114, 863–870.
Levanony H, Rubin R, Altschuler Y, Galili G. 1992. Evidence
for a novel route of wheat storage proteins to vacuoles.
Journal of Cell Biology 119, 1117–1128.
Lippincott-Schwartz J, Smith CL. 1997. Insights into secretory
and endocytic membrane traffic using green fluorescent
protein chimeras. Current Opinion in Neurobiology 7, 631–639.
Matsuoka K, Bassham DC, Raikhel NV, Nakamura K. 1995.
Different sensitivity to wortmannin of two vacuolar sorting
signals indicates the presence of distinct sorting machineries
in tobacco cells. Journal of Cell Biology 130, 1307–1318.
Matsuoka K, Nakamura K. 1991. Propeptide of a precursor to
a plant vacuolar protein required for vacuolar targeting.
Proceedings of the National Academy of Sciences, USA
88, 834–838.
Melchers LS, Sela-Buurlage MB, Vloemans SA, Woloshuk CP,
van Roekel JSC, Pen J, van den Elzen PJM, Cornelissen
BJC. 1993. Extracellular targeting of the vacuolar tobacco
proteins AP24, chitinase and b-1,3-glucanase in transgenic
plants. Plant Molecular Biology 21, 583–593.
Nagahora H, Ishikawa K, Niwa Y, Muraki M, Jigami Y. 1992.
Expression and secretion of wheat germ agglutinin by
Saccharomyces cerevisiae. European Journal of Biochemistry
210, 989–997.
Nakamura K, Matsuoka K, Mukumoto F, Watanabe N. 1993.
Processing and transport to the vacuole of a precursor to
sweet potato sporamin in transformed tobacco cell line BY-2.
Journal of Experimental Botany 44, 331–338.
Neuhaus J-M, Pietrzak M, Boller T. 1994. Mutation analysis
of the C-terminal vacuolar targeting peptide of tobacco
chitinase: low specificity of the sorting system, and gradual
transition between intracellular retention and secretion into
the extracellular space. The Plant Journal 5, 45–54.
Neuhaus J-M, Rogers JC. 1998. Sorting of proteins to vacuoles
in plant cells. Plant Molecular Biology 38, 127–144.
Neuhaus J-M, Sticher L, Meins Jr F, Boller T. 1991. A short
C-terminal sequence is necessary and sufficient for the
targeting of chitinases to the plant vacuole. Proceedings of
the National Academy of Sciences, USA 88, 10 362–10 366.
Okamoto T, Nakayama H, Seta K, Isobe T, Minamikawa T.
1994. Post-translational processing of a carboxy-terminal
propeptide containing a KDEL sequence of plant vacuolar
cysteine endopeptidase (SH-EP). FEBS Letters 351, 31–34.
Paris N, Rogers SW, Jiang L, Kirsch T, Beevers L, Phillips TE,
Rogers JC. 1997. Molecular cloning and further characterization of a probable plant vacuolar sorting receptor. Plant
Physiology 115, 29–39.
Paris N, Stanley CM, Jones RL, Rogers JC. 1996. Plant cells
contain two functionally distinct vacuolar compartments.
Cell 85, 563–572.
Roberts CJ, Nothwehr SF, Stevens TH. 1992. Membrane protein
sorting in the yeast secretory pathway: Evidence that the
vacuole may be the default compartment. Journal of Cell
Biology 119, 69–83.
Robinson DG, Bäumer M, Hinz G, Hohl I. 1998. Vesicle transfer
of storage proteins to the vacuole: the role of the Golgi
apparatus and multivesicular bodies. Journal of Plant
Physiology 152, 659–667.
Rothman JE, Wieland FT. 1996. Protein sorting by transport
vesicles. Science 272, 227–234.
Saalbach G, Jung R, Kunze G, Saalbach I, Adler K, Müntz K.
1991. Different legumin protein domains act as vacuolar
targeting signals. The Plant Cell 3, 695–708.
Saalbach G, Rosso M, Schumann U. 1996. The vacuolar
targeting signal of the 2S albumin from Brazil nut resides at
Cis-elements of protein transport
the C-terminus and involves the C-terminal propeptide as an
essential element. Plant Physiology 112, 975–985.
Saras J, Heldin CH. 1996. PDZ domains bind carboxy-terminal
sequences of target proteins. Trends in Biochemical Sciences
21, 455–458.
Schroeder MR, Borkhsenious ON, Matsuoka K, Nakamura K,
Raikhel NV. 1993. Colocalization of barley lectin and
sporamin in vacuoles of transgenic tobacco plants. Plant
Physiology 101, 451–458.
Scott MP, Jung R, Müntz K, Nielsen NC. 1992. A protease
responsible for post-translational cleavage of a conserved
Asn-Gly linkage in glycinin, the major seed storage protein
of soybean. Proceedings of the National Academy of Sciences,
USA 89, 658–662.
Shimada T, Kuroyanagi M, Nishimura M, Hara-Nishimura I
1997. A pumpkin 72-kDa membrane protein of precursoraccumulating vesicles has characteristics of a vacuolar sorting
receptor. Plant and Cell Physiology 38, 1414–1420.
Sonnewald U, von Schaewen A, Willmitzer L. 1990. Expression
of mutant patatin protein in transgenic tobacco plants: role
of glycans and intracellular location. The Plant Cell 2,
345–355.
Sticher L, Hinz U, Meyer AD, Meins F. 1992. Intracellular
transport and processing of a tobacco vacuolar
b-1,3-glucanase. Planta 188, 559–565.
Törmäkangas K. 1997. Structure, expression and intracellular
targeting of barley aspartic proteinase. PhD thesis, University
of Helsinki.
Vitale A, Bielli A, Ceriotti A. 1995. The binding protein
associates with monomeric phaseolin. Plant Physiology 107,
1411–1418.
Vitale A, Chrispeels MJ. 1992. Sorting of proteins to the
vacuoles of plant cells. Bioessays, 14, 151–160.
von Heijne G. 1990. The signal peptide. Journal of Membrane
Biology 115, 195–201.
von Schaewen A, Chrispeels MJ. 1993. Identification of vacuolar
sorting information in phytohemagglutinin, an unprocessed
vacuolar protein. Journal of Experimental Botany 44, 339–342.
Wilcox CA, Redding K, Wright R, Fuller RS. 1992. Mutation
of a tyrosine localization signal in the cytosolic tail of yeast
Kex2 protease disrupts Golgi retention and results in default
transport to the vacuole. Molecular Biology of the Cell 3,
1353–1371.
Wilkins TA, Bednarek SY, Raikhel NV. 1990. Role of propeptide
glycan in post-translational processing and transport of
barley lectin to vacuoles in transgenic tobacco. The Plant
Cell 2, 301–313.
Wink M. 1993. The plant vacuole—a multifunctional compartment. Journal of Experimental Botany 44, 231–246.
Young NM, Watson DC, Yaguchi M, Adar R, Arango R,
Rodriguez-Arango E, Sharon N, Blay PKS, Thibault P. 1995.
C-terminal post-translational proteolysis of plant lectins and
their recombinant forms expressed in Escherichia coli—
characterization of ‘ragged ends’ by mass spectrometry.
Journal of Biological Chemistry 270, 2563–2570.
KUNITZ-TYPE PROTEINASE INHIBITORS
Sporamin
Potato
Potato
Potato
Potato
Potato
...RFNPIRLPTTHEPA|...
...SQNLIDNLPS|ESPV..................VNENPLDVLFQEV-cooh
...SQNLIDLPS|ESPL..................VNENPLDVLFQEV-cooh
...SQNPINL|PSDATP..................VKDNPLDVSFKQVQ-cooh
...SQNPINL|PSDATP..................VKDNPLDVSFKQVQ-cooh
...SENPIVLPTTCHDDDN|LVLPEVYDADGNPLRIGER...
PAPAIN-LIKE CYSTEINE PROTEASES
Aleurain
...SSSSFADSNPIRPVTDRAAST...
Oryzain c
...ASSGFdDSNPIRSVTDHAASA...
Petunia
...RTANFADENPIRQVVSDSFHE...
Tomato
...GPATFADKNPIRQVVFDLPDE...
Arabidopsis
...SDVNDGDDLVIRQVVGGAEPQ...
Pea
...TDDTNNDDFIIRQVVDNEEDH...
2S ALBUMINS
Pumpkin
Pumpkin
Castor bean
Rapeseed
Arabidopsis
Brazil nut
...GIEN|PWRREG...KARNLPSMCGIRP-QRCDF*
...RATNLPSVCRLSQ-RRCELRSSRW*
...RSDN|QERSLR...TAANLPSMCGVSP-TECRF*
...DMEN|PQGPQQ...TATHLPKVCNIPQVSVCPFQKTMPGPS|Y*
...DMEN|PQGQQQ...TAKHLPNVCDIPQVDVCPFNIPSFPS|FY*
...YQTM|PRRGME...LAENIPSRCNLSP-MRCPMGGS|IAGF*
CEREAL LECTINS
Rice
CYKGGDGMAAILANNQSVSFEGIIESVAELV
Barley
CDG|VFAEAIAANSTLVAE
Wheat
CDG|VFAEAITANSTLLQE
Wheat
CDA|VFAGAITANSTLLAE
Wheat
CDG|VFAEAIATNSTLLAE
CHITINASES
Maize
Maize
Pea
Alfalfa
Bean
Vigna
Avocado
FNN NSASLAGTAAHAEA
STA HPCWNRCSAEA
FGS SLPLSSILLDTVAAA
FGS SLSLSSLFLNSIDT
FGN SLLLSDLVTSQ
FGN SLLNLHPIV
FGV STNPLAASS
173
174
Matsuoka and Neuhaus
Arabidopsis
Cotton
Ulmus
Tomato
Potato
Potato
Tobacco
Grapevine
Poplar
Poplar
FVN GLLEAAI
FGN GVSVDSM
FGN GLLLDTM
FGN GLLVDIM
FGN ALLVDTL
FGN GLLVDTV
FGN|GLLVDTM
FGS GLLLDTI
FGY GLSGLKDTM
FEDN GLLKMVGTM
GLUCANASES
Bean
Pea
Alfalfa
N.plumbaginifolia
N.glutinosa
Tobacco
Tomato
Potato
Hevea
FG AQRMQRLLLMSSMQHIPLRVTCKLEPSSQSLL
FG GERRDGEIVEGDFNGTVSLKSDM
FG G-ERMGIVNGDFNATI-SLKSDM
FG VSG-SVETNATA-SLISEI
FG VSGRV-DSSVETNATA-SLISEM
FG|VSGGVWDSSVETNATA-SLVSEM
FG VSERVWDI-TNSTASSLTSEI
FG VSERVWDISAETNSTTSSLISEM
FG A-EKNWDISTEHNATILFLKSDM
OSMOTINS
Thaumatococcus
Tobacco
Tobacco
S.commersonnii
Tomato
Arabidopsis
CP TALELEDE
CPN GQAHP-NFPLEMP-GSDEVAK
CPY GSAHNETTNFPLEMPTSSTHEVAK
CPY GSTHNETTNFPLEMP-TSTLEVA
CPN GVADP-NFPLEMP-ASTDEVAK
CPR SRLGATGSHQLPIKMVTEEN
PR-4
Tobacco
Potato
Tomato
Hevea
Soybean
Arabidopsis
CGDN MNVLVSPVDKE
CGDN VNVPLLSVVDKE
CGDN VNVPLLSVVDRE
CGDS FNPLFSVMKSSVINE
CGNE LDLTKPLLSILDAP
CGNE LIGQPDSRNMLVSAIDRV
PR-1
Tobacco
Tomato
RPFG DLEEQPFDSKLELPTDV
NVYG DLEEQKPDFDSKLELPN